Draw a 3D graph with the size of the square matrix along one independent axis,
e.g., from 1 to 100, and the number of available threads, e.g., from 1 to 16, along
the other showing the ratio between the number of dot products computed by
the most and the least loaded thread for different approaches to parallelizing the
two outermost for loops of matrix multiplication illustrated in Fig. 3.6.
Already registered? Login
Not Account? Sign up
Enter your email address to reset your password
Back to Login? Click here